important Flashcards
Anova
looks at the differences in mean between 1 categorical and 1 continuous variable
levene’s test
needs to be insignificant. looks at variance of scores between groups is equal of dv. homogeneity
type 1 error
finding an effect that does not exist
p value is 5 %
if study repeated the chance of finding the same effect would be 95%
between scores anova
scores are uncorrelated
repeated measure anova
data consist of repeated measures or matched
factorial anova
more than one categorical variable
non parametric anova
dv is ranked or ordinal or assumptions are violated
omnibus test
to reduce type 1 error when there is 3 teams. alpha level becomes 1-(1-a)^2
pearsons correlation i
standardized index of the linear relationship between 2 continuous variables
homoscedacity
constant variance of the y scores given x and visa versa
cross validation
subsamples reduce type 1 error
bonferroni procedure
alpha : number of correlations -> brings down type 1 error
r^2
variance in y that can be explained by x
covariance
sum of deviation scores added up devided by N
simpsons paradox
2 groups are combined which leads to a different outcome
residual
observed - predicted value
grand mean centering
mean becomes intercept and mean wil always be zero
multiple R is r is B
correlation between observed y and predicted y
adding a third variable
spuriosity- mediation- moderation
partial correlation
specific part of the total correlation associated with x1 and not x2
suppressor
variable that correlates with other independent variables but not dependent. effect increases or reverses
R1y=o and R1y.2=o
no linear association
r1y=R1y.2
replication
